CN112947374A - Intelligent self-healing control method for electric propulsion of regional distribution ship - Google Patents
Intelligent self-healing control method for electric propulsion of regional distribution ship Download PDFInfo
- Publication number
- CN112947374A CN112947374A CN202110181383.7A CN202110181383A CN112947374A CN 112947374 A CN112947374 A CN 112947374A CN 202110181383 A CN202110181383 A CN 202110181383A CN 112947374 A CN112947374 A CN 112947374A
- Authority
- CN
- China
- Prior art keywords
- fault
- healing
- self
- control
- electric propulsion
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000003745 diagnosis Methods 0.000 claims abstract description 44
- 239000002245 particle Substances 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 16
- 238000013528 artificial neural network Methods 0.000 claims description 15
- 239000000427 antigen Substances 0.000 claims description 12
- 102000036639 antigens Human genes 0.000 claims description 12
- 108091007433 antigens Proteins 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 9
- 238000011217 control strategy Methods 0.000 claims description 7
- 229960005486 vaccine Drugs 0.000 claims description 7
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000000306 recurrent effect Effects 0.000 claims description 6
- 230000002265 prevention Effects 0.000 claims description 5
- 230000035772 mutation Effects 0.000 claims description 3
- 238000013486 operation strategy Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 13
- 230000036039 immunity Effects 0.000 description 11
- 230000001360 synchronised effect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 239000002131 composite material Substances 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000035876 healing Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0262—Confirmation of fault detection, e.g. extra checks to confirm that a failure has indeed occurred
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B2219/00—Program-control systems
- G05B2219/20—Pc systems
- G05B2219/24—Pc safety
- G05B2219/24065—Real time diagnostics
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention provides an intelligent self-healing control method for electric propulsion of a regional distribution ship, which comprises the following steps: collecting operation data and information of an electric propulsion subsystem of a regional power distribution ship; if the fault is judged to occur, extracting fault characteristics according to the operating data and the information, dividing fault types by a particle swarm algorithm, and obtaining a fault sample; calculating the characteristic index of each fault sample; generating population individual optimum and global optimum according to the fitness function, updating the position and the speed of the particles according to the optimum value to generate a new clustering mode, and calculating the fitness function of the new particles; judging whether the clustering result meets the requirement; if the population is not satisfied and the maximum iteration times are not reached, returning to the step of generating population individual optimum and global optimum according to the fitness function to continue the iteration, if so, outputting a fault clustering result, and ending the particle swarm algorithm; and according to fault diagnosis, making a corresponding self-healing strategy and calculating the parameter structure of the fault-tolerant controller by using an immune algorithm.
Description
Technical Field
The invention relates to the technical improvement field of ship propulsion self-healing control, in particular to an intelligent self-healing control method for electric propulsion of a regional distribution ship.
Background
Modern ships are developing towards large-scale, electrification and intellectualization, the automation degree of a ship electric propulsion system is higher and higher, the functions are more complete, and the structure is more and more complex. The high requirements of the electric propulsion system and the electric power system on reliability and safety are determined by the characteristics of movement on demand and isolation and no aid of the large-scale ocean mobile platform, and therefore a strong electric power system and a reliable electric propulsion control system need to be explored and constructed.
The ship electric propulsion self-healing control technology is inheritance and development of the traditional fault diagnosis and fault-tolerant control technology, and comprises functions protection of a propulsion system during a fault period emphasized by the traditional fault-tolerant control, prevention and early warning before system fault, and active function recovery and fault immunity of a non-fault module after fault processing.
The concept of self-healing is derived from the medical field, self-healing is defined as the fact that a system perceives the self state in a systematic theory, and the self state is properly adjusted to recover the property of a normal state under the condition of no human intervention, so that the concept of self-healing has the remarkable characteristics of spontaneity, independence, action persistence and the like. In order to enable the ship regional distribution electric propulsion system to obtain the self-healing capacity, the state of the system needs to be automatically estimated through an intelligent means on the basis of a strong propulsion control system, faults are detected and diagnosed, fault reasons are analyzed, and a corresponding self-healing strategy is formulated to guide execution of self-healing control, so that the system obtains the self-healing capacity, the faults are eliminated or the influence is reduced, and the immunity and the reliability of an electric propulsion ship main power system to the faults are improved.
As a premise of self-healing control, detection and diagnosis of faults are being processed gradually by using intelligent methods, typically, fault diagnosis methods based on data driving, such as neural networks, support vector machines, group intelligent algorithms, and the like, and intelligent fault diagnosis methods based on mechanism models, such as expert systems, fault influence analysis, state estimation, and the like. Currently, research on self-healing control of a propulsion system mainly focuses on an intelligent fault-tolerant control strategy, and is mainly realized by model reference adaptive control, model prediction control and pseudo-inverse control,H ∞and the fault-tolerant control methods such as robust control and the like are combined with the fault diagnosis result to select a proper preset fault-tolerant control strategy, so that the system can still obtain good performance during the fault period. Furthermore, to enhance propulsion system controlThe system can be used for realizing reliability and forming advanced PID control methods such as fuzzy PID, neural network PID and the like by combining an intelligent method on the basis of the traditional PI control, so that the system has larger safety margin.
The principle of the intelligent fault diagnosis method based on the mechanism model is clear and easy to understand, but the fault model is difficult to accurately establish, and the diagnosis result is influenced. The intelligent fault diagnosis based on data driving adopts an intelligent method to carry out reasoning diagnosis on fault reasons through fault data, but the diagnosis precision is related to the data quality, a large amount of effective fault data is needed to carry out learning training, and the effective data is difficult to obtain. Robust control is conservative under normal conditions, and rapidity and robustness of a system cannot be considered at the same time, so that the robust fault-tolerant controller needs to sacrifice the stability of partial performance under normal working conditions for replacing faults. The pseudo-inverse control method has the advantages of simplicity and instantaneity, but cannot ensure the stability of a closed-loop system in a fault state. In addition, the current fault-tolerant control excessively depends on the accuracy and the real-time performance of a fault diagnosis mechanism, and when false detection or missing detection occurs, the fault-tolerant control performance cannot be guaranteed. The nonlinear control method has certain robustness, but the algorithm is complex and is not easy to realize, and the algorithm still needs to be simplified. The traditional PI control method is simple in structure and wide in application, but the actual working condition of a ship propulsion system is complex, and ideal closed-loop control performance cannot be obtained under the disturbance of loads such as propellers and the like. Although the intelligent method can improve partial defects of the controller, the closed-loop dynamic quality of the controller is sensitive to the change of the PI gain, so that the good dynamic quality of the controller is low in redundancy.
Disclosure of Invention
In view of the above drawbacks of the prior art, an object of the present invention is to provide an intelligent self-healing control method for electric propulsion of a regional distribution vessel, which solves the problem of self-healing of faults of a propulsion subsystem in a regional distribution vessel electric propulsion system during a sailing process.
In order to achieve the above objects and other related objects, the present invention provides an intelligent self-healing control method for electric propulsion of a regional distribution vessel, including:
collecting operation data and information of an electric propulsion subsystem of a regional power distribution ship;
judging the system operation performance according to the operation data and information of the electric propulsion subsystem of the regional distribution ship;
if the fault is judged to occur, extracting fault characteristics according to the operating data and the information, dividing fault types by a particle swarm algorithm, and obtaining a fault sample;
calculating the characteristic index of each fault sample, randomly generating a plurality of fault clustering modes as initial particle swarms, and calculating the fitness function of each particle;
generating population individual optimum and global optimum according to the fitness function, updating the position and the speed of the particles according to the optimum value to generate a new clustering mode, and calculating the fitness function of the new particles;
judging whether the clustering result meets the requirement or not by adopting a minimum square error sum criterion;
if the maximum iteration frequency is not met or is not reached, returning to the step of generating population individual optimum and global optimum according to the fitness function, updating the position and the speed of the particles according to the optimum values to generate a new clustering mode and calculating the fitness function of the new particles, continuing iteration, if the maximum iteration frequency is met, outputting a fault clustering result, finishing the particle swarm algorithm, and performing fault cause analysis and fault positioning through the fault knowledge base of the regional distribution electric propulsion ship;
and according to fault diagnosis, making a corresponding self-healing strategy and calculating the parameter structure of the fault-tolerant controller by using an immune algorithm.
In one implementation, the method further comprises:
generating initial fault-tolerant controller parameters as initial antibodies through data in a memory base, and if the number of the antibodies is insufficient, completing the antibodies through randomly generating data;
judging the control performance of the initial parameters according to the control performance indexes of the corresponding self-healing decisions;
if the requirement is not met, the affinity of the initial antibody and the antigen is used as the fitness to calculate, characteristics in the affinity are extracted to be used as a vaccine and a vaccine library is updated, meanwhile, the initial antigen is subjected to immune crossing and mutation operation to generate a new antibody, the affinity of the new antibody and the antigen is calculated, if the affinity does not reach the standard or the number of iterations is insufficient, the step of 'judging the control performance of the initial parameter according to the control performance index of a corresponding self-healing decision', if the algorithm ending condition is met, the corresponding fault-tolerant controller parameter is used as the antibody to update the antibody library, the self-healing control instruction coordination and distribution unit is used for parameter adjustment, self-healing control is executed, operation state judgment is carried out, and if the fault characteristics are eliminated.
In one implementation, in the case that the system operation performance is determined to have no fault according to the operation data and information of the regional power distribution ship electric propulsion subsystem, the method further includes:
short-term prediction is carried out on the system through an Elman neural network, and the method comprises the following steps: inputting historical operation data of ship electric propulsion and dividing the preprocessed data into a training set for training an Elman neural network and a test set for testing the network; training an Elman recurrent neural network by using a training set, judging whether network training is finished or not by combining the mean square error between prediction data and actual data with iteration times, using the trained network for short-term prediction of the state of the electric propulsion system of the regional distribution ship after testing, and using a prediction result for self-healing decision support and adjustment of the structural parameters of the coordination and distribution controller of the operation strategy of the propulsion system;
and if the potential fault exists, adjusting the structural parameters of the controller by adopting the control performance index in the prevention control strategy according to the prediction result.
As described above, according to the intelligent self-healing control method for electric propulsion of the regional distribution ship provided by the embodiment of the invention, the intelligent method is adopted to design the electric propulsion self-healing control system for the regional distribution ship, aiming at the problem of fault self-healing control of the electric propulsion system of the regional distribution ship. The intelligent fault diagnosis unit exerts the characteristics of particle swarm algorithm evolutionary computation and swarm intelligence, generates swarm intelligence to guide fault clustering through cooperation and competition among the swarm, and compared with the traditional evolutionary algorithm, the intelligent fault diagnosis unit can avoid complex genetic operation and dynamically track the current fault clustering condition by means of the special swarm memory to optimize the clustering result. The intelligent decision unit is used as an upper computer of the self-healing control system, structural parameters of the fault-tolerant controller are calculated by referring to a biological immunity principle through an artificial immunity intelligent algorithm, a fault self-healing fault-tolerant control strategy is formed to implement fault self-healing on the ship electric propulsion subsystem, and meanwhile, the self-healing control system can generate immunity on faults after the self-healing fault-tolerant control is completed by means of a learning and memorizing mechanism of the immunity algorithm, so that the stability of the propulsion subsystem is further improved. In addition, the propulsion subsystem trend prediction module in the intelligent self-healing decision-making system can directly reflect the time sequence characteristics of the dynamic process of the ship electric propulsion system by means of the Elman recurrent neural network, analyze and research the data with obvious time sequence characteristics, such as the propulsion system faults, carry out short-term prediction on the system state, adjust the structural state of the controller, prevent potential faults and further improve the immunity of the system to the faults.
Drawings
Fig. 1 is a schematic structural diagram of an intelligent self-healing control method for electric propulsion of a regional distribution ship according to an embodiment of the present invention.
Fig. 2 is a specific application schematic diagram of the intelligent self-healing control method for electric propulsion of the regional distribution vessel according to the embodiment of the invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention.
Please refer to fig. 1-2. It should be noted that the drawings provided in the present embodiment are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
The self-healing control method adopts the idea of closed-loop supervisory control, takes the regional distribution ship electric propulsion self-healing decision system as an upper computer, supervises the running state of the propulsion system in real time through system real-time data and fault diagnosis results provided by a data acquisition and intelligent fault diagnosis unit, formulates a fault-tolerant strategy according to the running state, coordinates conflicts among self-healing targets of all systems by an intelligent coordination and distribution unit, distributes self-healing tasks at the same time, adjusts the structural parameters of an electric propulsion fault-tolerant controller, and completes the self-healing control of the system. The intelligent fault diagnosis unit adopts a particle cluster intelligent clustering method to perform unsupervised clustering on the propulsion system faults to realize fault type identification and positioning, the intelligent decision unit adopts an artificial immune algorithm to calculate the structural parameters of the fault-tolerant controller according to the fault diagnosis result to make a corresponding self-healing strategy, and the intelligent coordination distribution unit adjusts the structural parameters of the self-healing fault-tolerant controller to realize fault-tolerant self-healing control under the fault state of the propulsion subsystem. When the system detects an obvious fault, the intelligent decision unit predicts the running state of the system in a short term through the Elman recurrent neural network, prevents a potential fault and improves the immunity of the system to the fault.
Fig. 1 is a structural block diagram of an electric propulsion self-healing control system of a large regional distribution ship, and the whole system consists of a ship regional distribution electric power system, an electric propulsion and control system, a data acquisition and fault diagnosis unit, an intelligent decision and coordination distribution unit and other ship subsystems.
1. Regional distribution ship electric propulsion system composition and control mode thereof
The self-healing control object area power distribution ship electric propulsion system comprises a ship area power distribution electric power system and a port and starboard propulsion unit. The ship regional power distribution power system is composed of four groups of high-power gas turbine generator sets and two groups of diesel generator sets in a regional power distribution grid structure, and provides stable and reliable power for the electric propulsion unit and other electric loads of ships. The propulsion unit consists of a transformer, a high-power frequency converter and a high-power permanent magnet synchronous propulsion motor. The instruction processing unit sends a control instruction of the propulsion motor according to a driving control trolley clock instruction, and the control instruction is executed by the high-power variable-frequency control unit to control the high-power propulsion motor to drive the propeller to rotate so as to form thrust to drive the ship to move.
In order to cope with a severe marine environment, a motor control system of a ship propulsion system generally includes two circuits of a rotation speed control system and a load control system, and is switched and determined by a selector switch. In the rotating speed mode control process, the control unit forms a rotating speed deviation loop to control the rotating speed of the motor according to the rotating speed feedback of the motor; when the system is in a load control mode in the case of heavy wind, waves and sea conditions, the torque load of the motor changes violently, the rotating speed control loop and the torque control loop run simultaneously in order to avoid the flying of the propeller, the rotating speed/torque signals are fed back to the frequency converter and the control unit of the frequency converter by the sensor, and the control distribution unit controls the propulsion motor in real time by referring to the safety limit of the ship. In general, the control method adopts PI control; if the frequency converter, the propulsion motor, the sensor and the like have faults, a corresponding self-healing strategy needs to be formed according to the fault diagnosis result, different fault-tolerant control modes are selected according to the self-healing decision to implement control and optimize control parameters, and self-healing control is implemented on the propulsion system under the faults.
2. Data acquisition and fault diagnosis unit of ship electric propulsion subsystem and operation process thereof
The unit collects electric data of the propulsion motor in real time through sensors of torque/rotating speed, voltage, current, power, frequency and the like and transmits the electric data to the control unit and the intelligent decision unit, and the fault diagnosis unit carries out fault diagnosis through an intelligent algorithm according to the collected data and transmits a diagnosis result to the intelligent decision unit to provide support for self-healing decision generation.
3. Intelligent self-healing decision and task coordination distribution unit for ship
The intelligent self-healing decision unit receives data of the data acquisition and fault diagnosis unit and performs global fusion with data of other subsystems, comprehensively evaluates the system, judges the state of the system, identifies a fault model according to a fault diagnosis result and fault data if the system fails, determines fault-tolerant control structure parameters by adopting an intelligent algorithm to form a self-healing strategy so as to ensure that the propulsion subsystems stably and effectively operate for a long time as a basic criterion, coordinates self-healing targets among the subsystems by the task coordination and distribution unit, distributes self-healing tasks, adjusts the structure parameters of the controller and performs fault-tolerant self-healing control on the fault propulsion system; if the propulsion system is in a normal state, the prediction module predicts the operation trend of the system by adopting an intelligent prediction method, prevents potential faults or selects a more economic operation mode according to navigation requirements.
Fig. 2 is a flow chart of a self-healing control method for electric propulsion of a regional distribution ship, wherein the self-healing control process is performed according to the sequence of firstly performing fault diagnosis, then determining a self-healing strategy and finally performing fault-tolerant self-healing control. The fault knowledge base comprises fault data such as a propulsion system sensor fault, a propulsion motor fault, a power supply fault and the like, wherein initial parameters are preset by a particle swarm clustering analysis algorithm through system fault knowledge, fault clustering analysis is carried out on collected data, a fault diagnosis result is provided for an intelligent decision unit, the decision unit calculates structural parameters of a fault-tolerant controller by adopting an artificial immune algorithm to generate a self-healing strategy, a coordination and distribution unit coordinates self-healing targets among systems to distribute control tasks, structural parameters of an electric propulsion fault-tolerant control unit are adjusted, and self-healing control is carried out on a fault system. The intelligent fault diagnosis comprises three processes of fault feature extraction, fault clustering and fault identification, and the fault identification can be used for respectively identifying faults of a propulsion system frequency converter, a propulsion motor, a sensor and a control system or identifying combined faults of all components. The intelligent self-healing decision and the coordination distribution comprise four links of state evaluation, fault model identification, self-healing fault-tolerant control decision and the coordination distribution. The fault-tolerant decision-making method comprises decision-making modes of online reconstruction, degraded operation, fault compensation, fault switching and the like of a control system, structural parameters of the fault-tolerant controller are optimized through an immune algorithm flow and are adjusted through a control distributor, and fault-tolerant self-healing control of the fault system is completed. If the system does not have obvious faults, an Elman recurrent neural network prediction algorithm is adopted to predict the propulsion system, parameters of the fault-tolerant controller are adjusted according to a prediction result and an immune algorithm flow, and potential faults are prevented or the economy of the system is improved.
Fig. 1-2 show an intelligent self-healing control method for electric propulsion of a regional distribution ship according to an embodiment of the present invention.
Fig. 1 is a block diagram of the overall structure of an electric propulsion self-healing control system of a regional distribution vessel, and 001-. Wherein 001 is a ship regional distribution power system, and is composed of four groups of high-power gas turbine generator sets 002, two groups of standby diesel generator sets 003 and a 004 port-board transformer 004 starboard transformer 005 so as to meet the power supply requirement of a large-scale electric propulsion ship port-board propulsion device.
The rotation speed torque sensor 006, the port high-power permanent magnet synchronous propulsion motor 007 and the port high-power frequency converter 008 form a port electric propulsion device, the rotation speed torque sensor 009, the starboard high-power permanent magnet synchronous propulsion motor 010 and the starboard high-power frequency converter 011 form a starboard electric propulsion device, and the two sets of propulsion devices can cooperate with each other, provide thrust for ships, can also be redundant for physical faults, and provide conditions for self-healing control.
The motor control mode of the ship propulsion system comprises a rotating speed control mode and a load control mode which are determined by switching of a selector switch, and the control method generally adopts PI control. The self-healing control process of the propulsion system of the regional power distribution ship is as follows: the instruction selection unit 012 generates a given rotating speed or torque according to the clock instruction, and in the rotating speed mode control process, the rotating speed control unit forms a rotating speed deviation loop with the given rotating speed to control the rotating speed of the motor according to the feedback information of the rotating speed sensors 006 and 009 of the port propulsion motor 007 and the starboard propulsion motor 010, which is acquired by the data acquisition unit 013; when the system is in a heavy wave and sea condition, the system works in a load control mode, the torque load of the motor changes violently at the moment, in order to avoid the propeller from flying, the rotating speed control loop and the torque control loop operate simultaneously, the rotating speed sensor 006 and the rotating speed sensor 009 input signals into the data acquisition unit 013 to perform torque calculation and then feed back the calculated torque to the fault-tolerant control unit 017, and the rotating speed and the torque of the propulsion motor are controlled by referring to the safety limit of the ship. The data collection and fault diagnosis module 013 is not only responsible for collecting sensor data to form a closed-loop control circuit, but also has the capability of monitoring the operation state of each device (unit 008-. If the propulsion system has faults, the intelligent fault diagnosis unit 013 conducts fault diagnosis according to three processes of fault feature extraction, fault clustering and fault identification, and the diagnosis result can provide support for self-healing decision. In the diagnosis process, fault recognition can be carried out on the port high-power frequency converter 008 and the starboard high-power frequency converter 011 on both sides, the port high-power permanent magnet synchronous propulsion motor 007 and the starboard high-power permanent magnet synchronous propulsion motor 010 in the propulsion system, the rotating speed sensor 006 and the rotating speed sensor 009 through the intelligent fault diagnosis unit 013 according to data acquired in real time, or fault recognition of combination fault energy of all components is carried out, so that the specific position of the fault is determined.
The self-healing decision unit 014 integrates the operation states of other subsystems 017 of the regional distribution electric propulsion ship according to the fault diagnosis result of the data acquisition and fault diagnosis module to select a proper self-healing decision, such as degraded operation or online reconstruction of serious faults, fault switching or fault compensation of faults caused by device failure, meanwhile, the parameters of the compound fault-tolerant control unit 016 are calculated by an immune optimization algorithm according to the identified fault model, the intelligent coordination distribution unit 015 controls the distribution unit to select a composite fault-tolerant control unit 016 structure (such as a PI control structure with mature control technology, a sliding-mode self-adaptive fault-tolerant control structure with stronger robustness and adaptability and a model prediction control structure with fault prevention capability) and adjust parameters thereof, so as to implement a corresponding self-healing control strategy and complete self-healing fault-tolerant control of the ship electric propulsion system.
The intelligent coordination and allocation unit 015 is responsible for allocating self-healing tasks to the subsystems according to self-healing decisions, and if conflicts occur between self-healing decisions of different subsystems, the unit performs policy coordination between the subsystems in the aim of preferentially ensuring the operation of the propulsion system. If the propulsion system is in a normal running state, the intelligent decision unit 014 predicts the failure of the propulsion subsystem and prevents the potential failure. The intelligent coordination distribution unit 015 adjusts the corresponding control modes of the two sets of propulsion units according to the self-healing decision made by the intelligent decision unit 014 by integrating the system operation states of other subsystems 017 of the regional distribution electric propulsion ship, thereby preventing faults and completing self-healing control.
Fig. 2 is a flow chart of a self-healing control method of an electric propulsion system of a regional distribution ship, wherein steps 2-11 are completed by a data acquisition and fault diagnosis unit, steps 12-18 and 21-27 are completed by an intelligent decision unit, steps 19,28 and 29 are completed by an intelligent coordination distribution unit, and steps 20 and 30 are implemented by a composite fault-tolerant controller.
According to the method, a step 2 of judging the system operation performance is carried out through collected operation data and information of a regional power distribution ship electric propulsion subsystem, a step 3 of extracting fault characteristics is carried out if a fault is judged, fault types are divided through a particle swarm clustering algorithm, and the specific process is that the extracted fault characteristics and historical data in a fault library are input into a fault clustering module and a fault clustering cluster is determined. Calculating characteristic indexes of each fault sample, randomly generating M fault clustering modes as initial particle swarms, calculating a fitness function 5 of each particle and generating population individual optimality and global optimality according to the characteristic indexes, updating particle positions and speeds according to optimal values to generate a new clustering mode and calculating a new fitness function step 7, judging whether a clustering result meets requirements or not by adopting a least square error sum criterion, if not, returning to the step 7 to continue iteration, if so, outputting a fault clustering result, finishing a particle swarm algorithm, performing a fault reason analysis step 9 and a fault positioning step 10 by using a fault knowledge base in the step 6, establishing a corresponding self-healing strategy step 12 according to a fault diagnosis result step 11, calculating a parameter structure of a fault-tolerant controller by using an immune algorithm, and performing self-healing by reconstructing a control system structure or properly reducing a control system performance index aiming at serious or emergent propulsion system faults And (4) healing control, the failure self-healing of the propulsion system can be implemented through switching or compensation control of the controller aiming at the failure faults of devices such as a controller, a high-power frequency converter, a sensor and the like.
The basic idea of the artificial immune algorithm is that an antigen in biological immunity corresponds to a target function and a constraint condition, an antibody corresponds to a solution vector in a space, and the solution of a problem is evaluated and selected by using the affinity between the antibody and the antigen, wherein the specific process comprises the steps of firstly identifying the antigen according to a fault model obtained by identification, generating initial fault-tolerant controller parameters as an initial antibody through data in a memory bank step 14, and if the number of the antibody is insufficient, completing the step 23 through randomly generating data. Judging the control performance of initial parameters according to the control performance indexes of corresponding self-healing decisions 15, if the requirements are not met, calculating the affinity of an initial antibody and an antigen as the fitness and extracting the characteristics as a vaccine and updating a vaccine library, meanwhile, generating a new antibody by the initial antigen through immune crossover and mutation 16, calculating the affinity of the new antibody and the antigen, if the affinity is not up to the standard or the iteration frequency is not enough 17, returning to 15, if the algorithm end conditions are met, updating the antibody library by using the set of fault-tolerant controller parameters as the antibody and outputting 18, performing parameter adjustment by a self-healing control instruction coordination distribution unit 19, executing 20, judging the operation state, if the fault characteristics are eliminated, performing short-term prediction on the system through an Elman neural network, and specifically, inputting the historical operation data of ship electric propulsion and dividing the data after the preprocessing 21 into data for training the Elman neural network A training set step 22 and a test set step 23 for testing the network. And (5) training the Elman recurrent neural network by using a training set, namely step 24, and judging whether the network training is finished in step 25 by combining the mean square error between the predicted data and the actual data with the iteration times. After the test, the trained network is used for short-term prediction of the state of the regional distribution ship electric propulsion system, the prediction result step 26 is used for self-healing decision support, the propulsion system operation strategy adjusting step 27 is used for coordinating and distributing structural parameters of the controller, and for a potential fault, the parameters of the composite controller can be adjusted by adopting a control performance index in a prevention control strategy according to the prediction result, the original control law or the controller structure is changed to improve the robustness of the system control margin enhancement system, the immunity of the potential fault enhancement system to the fault is prevented, or the controller performance is coordinated and optimized according to economic requirements, and the number of the propellers and the number of the generator sets are reasonably distributed so as to improve the sailing economy step 30.
The intelligent self-healing decision generation system provides support for the intelligent decision system self-healing decision generation through an intelligent fault diagnosis technology, and guides the advanced intelligent control unit to complete self-healing control of the regional distribution ship electric propulsion system. The method is particularly characterized in that an unsupervised particle swarm clustering method is adopted, fault types are divided through similarity of data characteristics to complete fault diagnosis, the clustering effect can be improved subsequently according to fault data in the running process of a ship, and the effects of saving a priori data training link and reducing the initial fault data demand are achieved. The intelligent decision unit can independently generate a self-healing decision by adopting an artificial immune algorithm according to the operation data of the ship electric propulsion system and the intelligent fault diagnosis result, and can record fault characteristics and corresponding decisions by means of the memory capacity of the algorithm to automatically generate a vaccine to improve the immunity of the system to the fault. In addition, the time sequence characteristic of the Elman neural network is used for short-term prediction of the ship electric propulsion system, potential faults can be prevented, navigation safety can be guaranteed, the prediction result can be used as a basis, and the running state of the propulsion system can be adjusted from an economic navigation angle.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (3)
1. An intelligent self-healing control method for electric propulsion of regional distribution ships is characterized by comprising the following steps:
collecting operation data and information of an electric propulsion subsystem of a regional power distribution ship;
judging the system operation performance according to the operation data and information of the electric propulsion subsystem of the regional distribution ship;
if the fault is judged to occur, extracting fault characteristics according to the operating data and the information, dividing fault types by a particle swarm algorithm, and obtaining a fault sample;
calculating the characteristic index of each fault sample, randomly generating a plurality of fault clustering modes as initial particle swarms, and calculating the fitness function of each particle;
generating population individual optimum and global optimum according to the fitness function, updating the position and the speed of the particles according to the optimum value to generate a new clustering mode, and calculating the fitness function of the new particles;
judging whether the clustering result meets the requirement or not by adopting a minimum square error sum criterion;
if the maximum iteration frequency is not met or is not reached, returning to the step of generating population individual optimum and global optimum according to the fitness function, updating the position and the speed of the particles according to the optimum values to generate a new clustering mode and calculating the fitness function of the new particles, continuing iteration, if the maximum iteration frequency is met, outputting a fault clustering result, finishing the particle swarm algorithm, and performing fault cause analysis and fault positioning through the fault knowledge base of the regional distribution electric propulsion ship;
and according to fault diagnosis, making a corresponding self-healing strategy and calculating the parameter structure of the fault-tolerant controller by using an immune algorithm.
2. The intelligent self-healing control method for regional power distribution ship electric propulsion according to claim 1, further comprising:
generating initial fault-tolerant controller parameters as initial antibodies through data in a memory base, and if the number of the antibodies is insufficient, completing the antibodies through randomly generating data;
judging the control performance of the initial parameters according to the control performance indexes of the corresponding self-healing decisions;
if the requirement is not met, the affinity of the initial antibody and the antigen is used as the fitness to calculate, characteristics in the affinity are extracted to be used as a vaccine and a vaccine library is updated, meanwhile, the initial antigen is subjected to immune crossing and mutation operation to generate a new antibody, the affinity of the new antibody and the antigen is calculated, if the affinity does not reach the standard or the number of iterations is insufficient, the step of 'judging the control performance of the initial parameter according to the control performance index of a corresponding self-healing decision', if the algorithm ending condition is met, the corresponding fault-tolerant controller parameter is used as the antibody to update the antibody library, the self-healing control instruction coordination and distribution unit is used for parameter adjustment, self-healing control is executed, operation state judgment is carried out, and if the fault characteristics are eliminated.
3. The intelligent self-healing control method for the electric propulsion of the regional distribution vessel according to claim 1 or 2, wherein when the system operation performance is determined according to the operation data and information of the regional distribution vessel electric propulsion subsystem without a fault, the method further comprises:
short-term prediction is carried out on the system through an Elman neural network, and the method comprises the following steps: inputting historical operation data of ship electric propulsion and dividing the preprocessed data into a training set for training an Elman neural network and a test set for testing the network; training an Elman recurrent neural network by using a training set, judging whether network training is finished or not by combining the mean square error between prediction data and actual data with iteration times, using the trained network for short-term prediction of the state of the electric propulsion system of the regional distribution ship after testing, and using a prediction result for self-healing decision support and adjustment of the structural parameters of the coordination and distribution controller of the operation strategy of the propulsion system;
and if the potential fault exists, adjusting the structural parameters of the controller by adopting the control performance index in the prevention control strategy according to the prediction result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110181383.7A CN112947374A (en) | 2021-02-09 | 2021-02-09 | Intelligent self-healing control method for electric propulsion of regional distribution ship |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110181383.7A CN112947374A (en) | 2021-02-09 | 2021-02-09 | Intelligent self-healing control method for electric propulsion of regional distribution ship |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112947374A true CN112947374A (en) | 2021-06-11 |
Family
ID=76245248
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110181383.7A Pending CN112947374A (en) | 2021-02-09 | 2021-02-09 | Intelligent self-healing control method for electric propulsion of regional distribution ship |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112947374A (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681442A (en) * | 2012-05-28 | 2012-09-19 | 上海海事大学 | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat |
| CN108334059A (en) * | 2018-02-26 | 2018-07-27 | 上海理工大学 | Based on particle cluster algorithm Optimized BP Neural Network Symbolic fault diagnosis method |
| CN108665112A (en) * | 2018-05-16 | 2018-10-16 | 东华大学 | Photovoltaic fault detection method based on Modified particle swarm optimization Elman networks |
| CN109376652A (en) * | 2018-10-24 | 2019-02-22 | 国网江苏省电力有限公司检修分公司 | Paralleling reactor of extra-high voltage method for diagnosing faults, apparatus and system based on artificial immune particle swarm clustering algorithm |
| CN110097236A (en) * | 2019-05-16 | 2019-08-06 | 南京工程学院 | A kind of short-term load forecasting method based on FA optimization Elman neural network |
| CN110263907A (en) * | 2019-07-01 | 2019-09-20 | 上海海事大学 | Based on the ship short trouble diagnostic method for improving GA-PSO-BP |
| CN110320467A (en) * | 2019-06-19 | 2019-10-11 | 武汉船用电力推进装置研究所（中国船舶重工集团公司第七一二研究所） | A kind of low-voltage direct circuit breaker failure diagnostic method |
| CN111351668A (en) * | 2020-01-14 | 2020-06-30 | 江苏科技大学 | Diesel engine fault diagnosis method based on optimized particle swarm algorithm and neural network |
| CN111815056A (en) * | 2020-07-10 | 2020-10-23 | 中国人民解放军空军工程大学 | Aircraft external field aircraft fuel system fault prediction method based on flight parameter data |
| CN112733962A (en) * | 2021-01-28 | 2021-04-30 | 上海海事大学 | Intelligent fault diagnosis and reconstruction method for ship regional power distribution power system |
| CN112736916A (en) * | 2020-12-29 | 2021-04-30 | 上海海事大学 | Immune multi-agent intelligent self-healing system and method for ship regional distribution power system |
| CN112819334A (en) * | 2021-02-03 | 2021-05-18 | 上海海事大学 | Intelligent fault-tolerant self-healing control method and system for power distribution of ship regional power distribution power system |
| CN112865301A (en) * | 2021-01-04 | 2021-05-28 | 上海海事大学 | Intelligent immune self-healing method and system for ship regional distribution power system |
| CN112865303A (en) * | 2021-01-06 | 2021-05-28 | 上海海事大学 | Self-sensing and self-diagnosing intelligent self-healing method for ship regional power distribution power system |
-
2021
- 2021-02-09 CN CN202110181383.7A patent/CN112947374A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681442A (en) * | 2012-05-28 | 2012-09-19 | 上海海事大学 | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat |
| CN108334059A (en) * | 2018-02-26 | 2018-07-27 | 上海理工大学 | Based on particle cluster algorithm Optimized BP Neural Network Symbolic fault diagnosis method |
| CN108665112A (en) * | 2018-05-16 | 2018-10-16 | 东华大学 | Photovoltaic fault detection method based on Modified particle swarm optimization Elman networks |
| CN109376652A (en) * | 2018-10-24 | 2019-02-22 | 国网江苏省电力有限公司检修分公司 | Paralleling reactor of extra-high voltage method for diagnosing faults, apparatus and system based on artificial immune particle swarm clustering algorithm |
| CN110097236A (en) * | 2019-05-16 | 2019-08-06 | 南京工程学院 | A kind of short-term load forecasting method based on FA optimization Elman neural network |
| CN110320467A (en) * | 2019-06-19 | 2019-10-11 | 武汉船用电力推进装置研究所（中国船舶重工集团公司第七一二研究所） | A kind of low-voltage direct circuit breaker failure diagnostic method |
| CN110263907A (en) * | 2019-07-01 | 2019-09-20 | 上海海事大学 | Based on the ship short trouble diagnostic method for improving GA-PSO-BP |
| CN111351668A (en) * | 2020-01-14 | 2020-06-30 | 江苏科技大学 | Diesel engine fault diagnosis method based on optimized particle swarm algorithm and neural network |
| CN111815056A (en) * | 2020-07-10 | 2020-10-23 | 中国人民解放军空军工程大学 | Aircraft external field aircraft fuel system fault prediction method based on flight parameter data |
| CN112736916A (en) * | 2020-12-29 | 2021-04-30 | 上海海事大学 | Immune multi-agent intelligent self-healing system and method for ship regional distribution power system |
| CN112865301A (en) * | 2021-01-04 | 2021-05-28 | 上海海事大学 | Intelligent immune self-healing method and system for ship regional distribution power system |
| CN112865303A (en) * | 2021-01-06 | 2021-05-28 | 上海海事大学 | Self-sensing and self-diagnosing intelligent self-healing method for ship regional power distribution power system |
| CN112733962A (en) * | 2021-01-28 | 2021-04-30 | 上海海事大学 | Intelligent fault diagnosis and reconstruction method for ship regional power distribution power system |
| CN112819334A (en) * | 2021-02-03 | 2021-05-18 | 上海海事大学 | Intelligent fault-tolerant self-healing control method and system for power distribution of ship regional power distribution power system |
Non-Patent Citations (10)
| Title |
|---|
| HOU WEIJUN 等: "Ship short term power load forecasting method based on support vector machine", SHIP SCIENCE AND TECHNOLOGY, 31 December 2018 (2018-12-31) * |
| 刘凯;李文佩;程潜善;方华亮;: "基于启发式规则和人工免疫的智能配电网自愈", 武汉大学学报(工学版), no. 02, 1 April 2016 (2016-04-01), pages 218 - 222 * |
| 张威;施伟锋;胡红钱;: "船舶区域配电系统全智能体重构及其优化研究", 电力系统保护与控制, no. 04, 3 February 2016 (2016-02-03) * |
| 梁瑞尤 等: "免疫原理的智能配电网自愈系统关键技术分析", 自动化与仪器仪表, 24 June 2020 (2020-06-24), pages 201 - 204 * |
| 王再英;: "基于人工免疫原理的系统异常检测与自适应容错控制", 化工自动化及仪表, no. 02, 10 April 2008 (2008-04-10) * |
| 胡红钱 等: "基于以太网的船舶电力系统动态电能质量监测与故障诊断系统设计", 中国舰船研究, 28 February 2018 (2018-02-28) * |
| 董欣;: "基于混合粒子滤波的故障诊断方法", 计算机系统应用, no. 12, 15 December 2012 (2012-12-15) * |
| 覃爱淞;张清华;胡勤;李铁鹰;: "人工免疫系统在智能故障诊断技术中的应用研究", 机床与液压, no. 19, 15 October 2013 (2013-10-15) * |
| 郑孝常 等: "船舶发电机组容错的StateFlow模糊PID控制研究", 控制理论与应用, 20 July 2015 (2015-07-20) * |
| 郑蕊蕊 等: "基于遗传支持向量机和灰色人工免疫算法的电力变压器故障诊断", 中国电机工程学报, 5 March 2011 (2011-03-05) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Review on deep learning applications in frequency analysis and control of modern power system | |
| US20220004191A1 (en) | Usv formation path-following method based on deep reinforcement learning | |
| Passino et al. | Introduction to intelligent control systems with high degrees of autonomy | |
| CN110110943A (en) | A kind of fleet's efficiency comprehensive intelligent Optimal Management System and optimization method based on big data | |
| CN102681442B (en) | Intelligent fault-tolerant control system and control method for zonal power distribution of full-electric boat | |
| CN109784201A (en) | AUV dynamic obstacle avoidance method based on four-dimensional risk assessment | |
| CN109407635B (en) | Intelligent system structure of comprehensive electric propulsion safety control system of ship | |
| CN116755474A (en) | Electric power line inspection method and system for unmanned aerial vehicle | |
| CN108563119A (en) | A kind of unmanned boat motion control method based on fuzzy support vector machine algorithm | |
| CN110994673B (en) | Prediction method for micro-grid self-adaptive anti-islanding disturbance load impedance value | |
| CN113268078A (en) | Target tracking and trapping method for self-adaptive environment of unmanned aerial vehicle group | |
| CN114578792A (en) | Multi-agent fault diagnosis method and system | |
| Fan et al. | Development trend and hotspot analysis of ship energy management | |
| CN114217519A (en) | Self-reconfigurable marine network control system architecture and self-reconfigurable method | |
| CN113141012A (en) | Power grid power flow regulation and control decision reasoning method based on deep deterministic strategy gradient network | |
| CN112327098A (en) | Power distribution network fault section positioning method based on low-voltage distribution network comprehensive monitoring unit | |
| CN113220034B (en) | Unmanned aerial vehicle cluster reconstruction system combining autonomous reconstruction and manual intervention reconstruction | |
| CN112947374A (en) | Intelligent self-healing control method for electric propulsion of regional distribution ship | |
| CN112230548A (en) | Nuclear power device autonomous control system | |
| CN112733962A (en) | Intelligent fault diagnosis and reconstruction method for ship regional power distribution power system | |
| CN116362377A (en) | Large power grid region cooperative power flow regulation and control method based on multi-agent strategy gradient model | |
| CN112865303A (en) | Self-sensing and self-diagnosing intelligent self-healing method for ship regional power distribution power system | |
| CN114298429A (en) | Power distribution network scheme aided decision-making method, system, device and storage medium | |
| CN112736916B (en) | Intelligent self-healing system and method for immune multi-agent of ship regional distribution power system | |
| CN112819334A (en) | Intelligent fault-tolerant self-healing control method and system for power distribution of ship regional power distribution power system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |